The Daily Brief

India's Voice AI Boom: Bajaj Finance and Axis Bank Scale AI Agents Despite Accuracy Hurdles

24 min Listen to the original Episode page
Voice AIIndian FinancialsBajaj FinanceAxis BankSarvam AICost EfficiencyData PrivacyLinguistic Diversity

Summary

The episode explores the rapid adoption of voice AI in India's financial sector, driven by linguistic diversity and cost efficiencies over human telecallers. Speakers detail how Bajaj Finance improved conversion rates from 8% to 17% after retraining models on internal data, generating 2,504 crore rupees in loans via voice bots between April and June 2026. Axis Bank utilized AI for 60% of collection calls in Q2 2026, retaining data to maintain vendor leverage. The discussion highlights technical challenges like latency in the STT-LLM-TTS relay and poor representation of Indian languages in global models like OpenAI's Whisper. Additional segments cover how fuel shortages benefited large tile manufacturers and brief updates on Hindalco, solar jobs, and AI chip export rules.

Episode sections

  1. 0:04 – 12:49

    Intro and India's Voice AI Boom

    The episode introduces the rapid growth of voice AI in India, noting that banks are already using it for sales and collections despite challenges in accuracy and profitability. It highlights the linguistic diversity of India as a key hurdle for AI models and sets up the discussion on startups and major financial institutions adopting the technology.

  2. 12:49 – 20:59

    How Factory Shutdowns Helped Tile Giants

    The second story explains how fuel shortages and the war in West Asia disrupted smaller Morbi tile manufacturers, leading to higher profits for large branded makers like Kajaria and Somany due to reduced competition and higher prices. It contrasts this with Cera's struggles due to rising brass costs and questions the sustainability of this advantage.

  3. The segment covers five quick news items: Hindalco's capacity expansion, a study on solar job creation, US export rules for AI chips via Southeast Asian data centers, India-Thailand trade growth, and India's push for greenfield shipbuilding clusters.

Assets discussed

AssetKindStanceContext
NVIDIANVDAequityneutralThe license would be required before exporting advanced AI chips from Nvidia. and other American manufacturers.

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

India has 22 scheduled languages written in different scripts, making typing difficult for native speakers.

Unidentified speaker (speaker_0) ·0:43

How to verify: Constitution of India (Eighth Schedule); linguistic surveys by the Ministry of Culture. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

Meta has not shared exact usage figures for voice notes, but surveys suggest India uses them more than most Asian countries.

Unidentified speaker (speaker_0) ·1:03

How to verify: Meta earnings reports or public statements; third-party surveys like Pew Research or local digital usage reports. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

OpenAI's Whisper original training data contained 680,000 hours of audio, with 438,000 hours of English speech.

Unidentified speaker (speaker_0) ·4:21

How to verify: OpenAI's Whisper technical paper or documentation detailing dataset composition. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

34 companies are currently building AI for India's languages, according to Taction data.

Unidentified speaker (speaker_0) ·5:00

How to verify: Taction market reports or press releases citing this statistic. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

Bajaj Finance states a voice AI agent costs about a third of what a human one does.

Unidentified speaker (speaker_0) ·5:40

How to verify: Bajaj Finance investor presentations, earnings call transcripts, or press releases. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

Human telecallers in India talk for 90-120 minutes a day, with 30-60% attrition annually.

Unidentified speaker (speaker_0) ·5:40

How to verify: Industry reports from HR consultancies or telecom sector studies. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

Sarvam's SARS version 3 model got about one word in five wrong in public tests against OpenAI Indic voices.

Unidentified speaker (speaker_0) ·7:25

How to verify: Sarvam public test results or benchmark reports published by the company. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

Bajaj Finance lost 640 crore rupees in potential business due to a 12% call drop rate in early AI adoption.

Unidentified speaker (speaker_0) ·8:45

How to verify: Bajaj Finance earnings call transcripts (April 2026) or investor presentations. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

Between April and June 2026, Bajaj Finance loans from voice and text bots totaled 2,551 crore rupees, with 2,104 crore from voice bots.

Unidentified speaker (speaker_0) ·9:44

How to verify: Bajaj Finance Q2 FY2026 earnings report or press release. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

Between July and September 2026, Bajaj Finance generated 596 crore rupees in loans from leads found via AI analysis of 56 million customer voice interactions.

Unidentified speaker (speaker_0) ·10:24

How to verify: Bajaj Finance Q3 FY2026 earnings report or press release. For recent events, verify 2026-08-18 through 2026-11-18. Use another period only when the speaker explicitly names it.

Do your own research

  • Verify Bajaj Finance's Q1-Q3 2026 quarterly filings for the reported 2,504 crore rupees in voice bot-generated loans and any disclosures regarding AI technology costs or risks.
  • Check Axis Bank's Q1 and Q2 2026 earnings reports for details on collection call volumes and any partnerships with AI vendors.
  • Review Sarvam AI's technical documentation or whitepapers for the reported Word Error Rate (WER) of one word in five on Indian languages.
  • Compare the cost per minute of AI voice agents (1.75-2 rupees) versus human telecallers (6-10 rupees) using industry reports from NASSCOM or similar bodies.
  • Investigate the specific terms of data retention agreements between banks like Axis Bank and AI vendors to assess privacy risks and vendor lock-in.
  • Analyze the impact of OpenAI's Whisper model language distribution on the performance of global AI models in non-English Indian languages.
  • Assess the sustainability of Kajaria and Somany's profit margins given the temporary nature of fuel shortage-driven competitive advantages in the tile industry.
  • Review US export control rules for AI chips and their potential impact on Southeast Asian data centers as mentioned in the tidbits section.
Full transcriptexpand
0:04
Unidentified speaker (speaker_0): India's voice AI industry is growing. Banks are already using AI to sell loans, collect payments, and speak to customers in Indian languages. Startups are racing to build these systems, but making them accurate, affordable, and profitable is proving difficult. That's our first story. Next, India's big
0:23
Unidentified speaker (speaker_0): tilemakers are reporting sharply higher profits after the war in West. Asia disrupted their smaller competitors. But can that advantage last? Welcome back to the daily brief by Zeroda, where we cut through the noise to help you understand what's actually happening in the most important stories from business and markets. I'm your host Sahiti, and today is
0:43
Unidentified speaker (speaker_0): Wednesday, November 18, 2026. Coming to our first story. For many Indians, using their voice online is much more than convenience. India has 22 scheduled languages written in different scripts from Dev Nagri to Tamil. Typing them is difficult even for native speakers and readers because they have
1:03
Unidentified speaker (speaker_0): to navigate different keyboards. Smartphones are still mostly designed for Kuwait. So you might prefer sending a WhatsApp voice note instead. Meta hasn't shared exact usage figures, but some surveys, such as suggest India uses voice notes more than most Asian countries. Until now, much of the AI built out has focused on chatbots and text, but voice
1:23
Unidentified speaker (speaker_0): AI is catching up fast and already changing some familiar routines. Have you ever called customer care and heard press 1 for Hindi, 2 for English, 3 for a human before waiting through painfully cheerful elevator music? Voice AI replaces that with. a phone conversation in Hindi, Tamil, and even a mix of both.
1:43
Unidentified speaker (speaker_0): Voice AI is already affecting business outcomes, especially at some of India's biggest banks and NBFCs like Access Bank and Bajaj Finance, and dozens of Indian startups want a piece of that market. Today we look at what it takes for AI to hold a phone call the way Indians actually speak. A quick note: we will be mostly be looking at
2:03
Unidentified speaker (speaker_0): AI that holds live phone conversations, not apps that transcribe meetings or dictation like granula or vesperflow. With ChatGPT, you type a question in a large language model or LLM generates an answer by predicting the next token. But voice AI has an extra challenge. It needs to understand what you're saying and respond quickly enough to feel
2:23
Unidentified speaker (speaker_0): like a human conversation. A standalone LLM cannot do that. effectively, so the most common approach uses three AI models in a relay. First, a speech-to-text model or STT turns someone's speech into written words. Next, a standard LLM like Opus 5.5 works out what the person meant, checks the context and its
2:43
Unidentified speaker (speaker_0): instructions, and writes a reply. Finally, a text-to-speech model or TTS turns that reply back into a voice. Simple enough, right? But using three AI models takes time. Human conversations have quick back and forth reactions. We barely notice a chatbot taking time to reply, but 3 seconds of silence on a phone call with
3:02
Unidentified speaker (speaker_0): a voice bot feels awkward. Every step in the relay adds delay, which the industry calls latency. There's another approach to voice AI, which isn't mainstream in India, yet, due to limited Indian language support and high costs. Instead of using three models, companies like OpenAI and Google have built just one direct speech to
3:22
Unidentified speaker (speaker_0): speech model to do all of the heavy lifting. Beyond latency, there's also the problem of understanding what is being spoken. You see, traditional telephone audio carries around 8,000 samples of sound every second. The STD model chops the stream into slice. slices a few milliseconds long. It turns each slice
3:42
Unidentified speaker (speaker_0): into a picture of which thousand frequencies are present and how loud they are. Then it learns which pattern represents which sound. For instance, in one test of Bengali speech published by a voice AI startup called Navana, competitor voice AI agents heard Rajanikanda, the word for the tuberos tubrose flower, as Rajanikan, Rajanikand, the legendary
4:02
Unidentified speaker (speaker_0): Tamil Air. actor. Imagine if on a loan collection call, 15,000 was confused with 50,000 rupees. This is where training data comes in. A speech model learns from recordings of people speaking paired with accurate transcripts. Take OpenAI's Whisper. Its original training data contained
4:21
Unidentified speaker (speaker_0): 6,80,000 hours of audio. Of these, 4,38,000 hours were English. speech paired with English transcripts, and another 1,26,000 hours involved translations. Only the remaining 1,17,000 hours had non-English speech paired with transcripts in the original language spread across
4:40
Unidentified speaker (speaker_0): 98 languages. That means Indian languages are poorly represented Presented in large datasets like these. In the text isn't enough either. Models need recordings of people actually speaking across different accents, dialects, and noisy environments, along with accurate transcripts. That's the enormous problem
5:00
Unidentified speaker (speaker_0): Indian startups have set out to solve. As per data from Taction, 34 companies are currently building AI for India's languages. Among those working on voice, a few like Sarvam, Nani.ai, and Nirvana built build their own species. speech models. Others like Bolna and Ring.ai build platforms that stitch other companies'
5:20
Unidentified speaker (speaker_0): models together into working voice agents. Their pitch to companies is simple a voice agent can make routine phone calls to sell loans or chase payments for less than a human. As per Bhajaj Finance, a voice AI agent costs about a third of what a human one does. How so? Voice startup Walkstone. Waxton estimates that a human
5:40
Unidentified speaker (speaker_0): telecaller actually talks for just 90 to 120 minutes a day. At the desk, supervisors end the cost of replacing the 30 to 60% of telecallers who quit every year. That same estimate puts each minute of human conversation at 6 to 10 rupees. By contrast, a large company buying AI calls in
5:59
Unidentified speaker (speaker_0): bulk can pay as little as 15 1.75 pesos. pesa to 2 a minute.
6:05
Unidentified speaker (UNKNOWN): You
6:05
Unidentified speaker (speaker_0): can buy cheaper enterprise plans or choose pay as you go plans. Anyone can sign up online and pay by the minute, much like the prepaid phone plan. These plans cost around 3 to 6 rupees per minute. Still, even at a fraction of the cost, the savings in India are smaller than in other countries because labor is already cheaper here. In the US,
6:25
Unidentified speaker (speaker_0): for comparison, a voice agent. can be 10 or more times cheaper than a person. Currently, that makes it harder for Indian voice AI providers to cut prices further. Meanwhile, startups also have high costs. For every minute of a call, they pay for both the phone line and the models underneath. The cost is higher for startups that rent most of
6:45
Unidentified speaker (speaker_0): their stack. Nani.aiCo Nani.ai CO Ganesh Kupalan says these aggregators struggle. to make money because each layer carries an outside cost. Building your own isn't cheap either. Ring started out as a TTS company but found training its own speech models too costly. That leaves quality as the main way to stand out.
7:05
Unidentified speaker (speaker_0): One measure of quality in voice AI is latency, which we just mentioned above. Accuracy is another measure. It is tracked using a metric called word error or WER, which is the share of words a model gets wrong. Global leaders mostly test WER for English, but for Indian languages, companies
7:25
Unidentified speaker (speaker_0): largely publish their own tests. Sarwam says its SARS version 3 model beat OpenAI Indic voices a public test of Indian speech while still getting about one word in five wrong. This makes it a hard for a buyer. to tell vendors apart and it comes back to price. Ring found basic outbound calling and loan
7:45
Unidentified speaker (speaker_0): collection were in particularly sticky businesses because they are already highly commoditized. Meanwhile, relationship-heavy work like wealth management still largely stays with humans. The call startups that are trusted with are also the easiest to move to another vendor. That's why the sector has, by some accounts, already started seeing price wars. One mode
8:05
Unidentified speaker (speaker_0): lies in integration. For a bank, a voice agent is only useful if it's wired into the bank system. It has to see a borrower's loan details, record a promise to pay, and send UPI payment link during the call, all the while following RBI's rules. Building all that takes months for each new customer, which is why these startups are selling
8:25
Unidentified speaker (speaker_0): more like service firms rather. than software companies. This is also where the biggest and first adopters of Voice AI in India, banks and NBFCs come into the picture. After all, they make a huge number of calls and each has an outcome they count, a loan sold or an EMI collected. Pajaj Finance, India's
8:45
Unidentified speaker (speaker_0): largest NBFC, has shared the most detail about how it's going. And its start was in wasn't small. smooth. Initially, their voice bots lost track of conversations after about 3 minutes, paused too long before replying, or struggled with background noise. About 12% of calls dropped, costing its potential business worth 640 crore rupees.
9:05
Unidentified speaker (speaker_0): But Bajaj kept working at it. Over 18 months, working with an unnamed technology partner, it's retrained its STT model in six languages on its own passcode. past calls. Its bots conversation rate has since roughly doubled from 8% to 17%. The AI bots now do most of the work that people used to do. On
9:24
Unidentified speaker (speaker_0): Bajaj April 2026 earnings call, managing director Rajev Chin said the company once had 5,000 human outbound calling agents. Now 30% of the agents it has left are AI. But we still cannot confirm exactly. how many human jobs and the bots have replaced at Bajaj. The bots are also being tied to
9:44
Unidentified speaker (speaker_0): material business's outcomes now. Between April and June 2026, loans from leads generated by Bajaj Voice and Textbots totaled 2,551 crore. Of that, 2104 crore rupees came through voice bots pots alone. Baj Bajaj doesn't say how much bots pods add to its earnings. but there's an
10:04
Unidentified speaker (speaker_0): important metric at the end of it. Bajaj also uses AI not merely to talk to customers but to listen to them. It uses AI to analyze recordings of customer calls, turning them into data that flags who might want to loan. Between July and September 2026 alone, it turned 56 million customer voice interactions into data set
10:24
Unidentified speaker (speaker_0): this way and loans from leads found this way amounted to 596 crore rupees. On the bank side, Access Banks runs its AI work under a bank-wide program called Axiom, which covers document reading, voice, and more. In the second quarter of 2026, VoiceBots led 60% of Access Bank collection calls and more
10:44
Unidentified speaker (speaker_0): than 2.5 million sales calls. In the first quarter, it's what made more than 80 million collection calls. But it hasn't said what any of this saves or earns. Whatever vendor powers the voice bot, Bajaj and Access retains the calls themselves. Those recordings and whatever can be learned from them usually sit with them,
11:04
Unidentified speaker (speaker_0): not the startups. But what about the voice AI market beyond Indian finance? How big is that really? Well, that's the problem. The work voice AI currently does is the repetitive low value kind. And there are only so many customers that are willing to spend money at scale to automate that work. As per two venture investors, only
11:24
Unidentified speaker (speaker_0): about 100 companies in India are large enough to sign big voice AI contracts, and that's mostly banks, insurers, and government services. Even those rarely hand all of their calls to one vendor to avoid dependency. And even among potential customers, adoption is very limited.
11:41
Unidentified speaker (UNKNOWN): A
11:41
Unidentified speaker (speaker_0): NASCOM survey of over 140 organizations found that only 13% had scale Voice AI across multiple customer journeys. Voice AI already has business outcomes we can measure. But so far, plenty of suppliers are chasing a much smaller pool of demand, even though it's growing. Startups have to compete on
12:01
Unidentified speaker (speaker_0): price, and enterprise. enterprises haven't yet handed over complicated tasks. The models can be rented and the vendors swapped, while ears of data with banks' own customers cannot be bought anywhere else. That's why at the top of the market, the stronger hand seems to sit with banks and NBFCs like Pajaj and Access. Meanwhile,
12:21
Unidentified speaker (speaker_0): the rules are catching up too. In September, Tri changed its regulations so that businesses might must tell their telecom operator in advance about any calls placed by software, including those using artificial voices. Undeclared calls will be treated as spam. The next time your bank calls about an EMI, there's a fair chance the voice on the line
12:41
Unidentified speaker (speaker_0): is AI. Soon the network will at least know which ones are.
12:49
Unidentified speaker (speaker_0): Coming to the second story.
12:51
Unidentified speaker (UNKNOWN): In
12:52
Unidentified speaker (speaker_0): March, Morbi ran short of fuel. It's a city in Gujarat, famously called the ceramics capital of India. The war in West Asia disrupted the supply of propane and natural gas that fire the cleansed there. By March 31st, only 83 of its roughly 800 ceramic units were still running. Yet, when the July to September quarter
13:11
Unidentified speaker (speaker_0): ended, India's big branded tile makers reported sharp. sharply higher profits. These two things are connected. Tiles are pressed from clay and baked in kilns for hours, so energy is one of the biggest costs in the business. Mordbi makes nearly 90% of India's ceramic tiles and sanitary wear. Branded companies like Kajaria
13:32
Unidentified speaker (speaker_0): and Somani compete with these factories, run their own plants elsewhere, and also buy tiles from Morbi to sell under their own. names. When we last wrote about this sector in September 2025, the problem was too many tiles. Morbi was making more than the market could absorb, which kept prices low and squeezed the big brands.
13:52
Unidentified speaker (speaker_0): This quarter looks like the reverse. So has the industry turned a corner or did a was wash simply hand a few big players an advantage they may not keep? Look at the latest results. and you'll see what we mean. Kajaria Ceremix, India's largest tile maker, grew revenue by about a quarter and net profit by more than 40%.
14:12
Unidentified speaker (speaker_0): Somani Ceramix grew revenue just as fast and its profits more than tripled. Neither got there by selling many more tiles. Somani sold only 5% more, partly because some plans were down for upgrades in July and August. So most of its growth came from higher prices. Kajaria sold 10% more, so for
14:32
Unidentified speaker (speaker_0): it, the split was roughly half volume, half price. Serra Sanitary Wear, which mostly sells toilets, washpasins, and tabs, is the odd one out. Its revenue grew even faster by 27%, yet its profit slipped slightly. It was only one of the three without a blockbuster quarter. because it cost
14:51
Unidentified speaker (speaker_0): rose on two friends at once. We will get to that later. So, all three sold more, but not all of them kept more of it. More than half of Morby's factories ran on propane before the war. And from March 9, the government also kept how much piped gas each industrial user could draw. Factories restarted through April on piped pipe gas,
15:11
Unidentified speaker (speaker_0): and by mid-May, more than 670 were running again. But they were running at a very different cost. Pipe gas in Morbi now costs nearly twice what it did before the war. Propane returned in July and many factories switched back to it, but by late October it had become almost as expensive. Coal isn't an option either, since
15:31
Unidentified speaker (speaker_0): 2019 National Green Tribunal order shed the coal gasifiers Morbi once used. Kajaria largely escaped the squeeze.
15:39
Unidentified speaker (UNKNOWN): It
15:39
Unidentified speaker (speaker_0): owns three plants in Morbi through subsidiaries and raised prices there by 18 to 19%. But most of its capacity is in the north and south, where it kept getting gas from gale at contract prices well below Morbi's. Its northern plants also run partly on biofuel. Somani sits somewhere in between. Its biggest plant is
15:59
Unidentified speaker (speaker_0): in Haryana, but most of its joint venture plants are in Morbi. For a sense of what Morbi's style cost did to profits, look at Gujarat-based Asian granito. Its revenue grew by 19%, but its power and fuel bill rose 53% and its profit fell about 75%.
16:17
Unidentified speaker (speaker_0): Morbi's smaller makers had even less protection, and Kajaria described what it did to their prices. The same war hit Morbi's demand side. Exports make up about a quarter of the industry sales. Between April and July, India's title exports fell 34% in value and 46% in square meters from a year earlier.
16:37
Unidentified speaker (speaker_0): April alone was down 72%. The steepest falls were in the Gulf. Exports to the UAE fell 62%, shipments to Kuwait quite almost vanished, and Iraq fell 95%. But the Gulf was only a small part. of India's style exports, so it cannot explain most of that slump. The war raised freight costs
16:56
Unidentified speaker (speaker_0): on most routes out of India. Shipping a container to the Middle East cost about 10 times more once the war broke out. Even rates to London rose by well over half. Shipments to the US fell 36% and were also under pressure from an anti-dumping case against Indian tiles. Only Russia grew slightly, making it
17:16
Unidentified speaker (speaker_0): India's largest tile market so far this year. Now you might expect all those unsold export tiles to flood India and drag prices down as they did last year. That hasn't really happened. With fuel this expensive, Morby's export-focused makers couldn't afford to sell cheaply at home, so many simply shut instead. Last year's
17:36
Unidentified speaker (speaker_0): problem was too many tiles in and the water. the ward changed that by taking supply away. That left the big brands with something years of competing with Morbi never got them. A smaller price gap. Morbi had to raise prices far more than the branded makers did. So Morbi tile is no longer that much cheap. With branded tiles, now a much smaller step up in price,
17:56
Unidentified speaker (speaker_0): Kajaria sold 10% more tiles in a market that by its own estimate grew only 5 to 6%.
18:03
Unidentified speaker (speaker_0): You would expect Serra, one of the biggest names selling into the same houses and building projects, to have benefited too, right? It didn't because tiles are only a small part of its business. Most of Serra sales come from bathroom products, and that's where cost hits hardest. For Kajaria and Somani, it's opposite. It, Bathroom products are only about a tenth
18:23
Unidentified speaker (speaker_0): of sales. Tabs, which make up over 40% of Serra Serra's sales. sales, are largely made of brass, and brass costs Serra 58% more than a year earlier. Its toilets and wash basins are ceramic too, so they also felt the squeeze to less than tilested. Sarah's own Gujarat plant paid more for gas, and it buys about 60% of its
18:43
Unidentified speaker (speaker_0): sanitary wear from outside makers, many of them in Marbi. Marby. Serra raised tap prices by about 8%, but only in September. too late to help this quarter. Its operating margin fell from 14.2% to 12.2%, and it now needs a much stronger second half to meet its full-year target. Somani, which also
19:03
Unidentified speaker (speaker_0): sells tabs, says the whole industry has struggled to pass on higher brass costs. It also says pricing follows the market leader, Jaguar. The brand in inmakers makers are petting their advantage will last. Kajaria isn't raising prices further. Instead, it's using the smaller price gap to win buyers. Both companies are expanding plants
19:22
Unidentified speaker (speaker_0): they already own. Kajaria new lines in Andhra Pradesh and Rajasthan are due by mid-2027. After that, it plans to buy fewer tiles from Morbi. That's a reversal for brands that spent years outsourcing production to Morbi because it was cheaper. Somani's joint venture in the south is also doubling its planned. Both
19:42
Unidentified speaker (speaker_0): companies know this year's price boost won't repeat. So Kajaria is focusing more on tile adhesive, which builders use more as tiles get bigger. It expects adhesive sales to grow from about 125 crore last year to roughly 400 crore next year. By early November, Mordbi's Mod B's gas prices had already eased a
20:02
Unidentified speaker (speaker_0): little. But Somani doesn't expect fuel prices to return to pre-war levels even if the war ends. That is a forecast and the risks run other way. If fuel becomes cheaper, Mod B can quickly reverse its price increases and the narrower price gap disappears. The shut factories are also export units waiting for orders. If they restart before exports
20:22
Unidentified speaker (speaker_0): recover, their tiles will end up in India. Freight to the Gulf was still near a 5-year high in late October. But Somani said in early November that Mordbees export had improved over the past month as freight eased, with a proper restart expected only in January after Europe's December shutdown. What's done is done.
20:42
Unidentified speaker (speaker_0): The real test comes when Mordbees costs fall, and we see whether buyers who switch to branded tiles deal with them. And if they do, the war will have achieved. something for the big brands that years of advertising could not.
20:59
Unidentified speaker (speaker_0): And now, a few quick tidbits. Number one: Hindustan Zinks Board has approved an investment of about 80,000 crore rupees over the next 3 to 5 years. It plans to increase its metal production capacity from 1.1 million tons per year. to 2 million tons per year. The company expects to reach
21:18
Unidentified speaker (speaker_0): 1.4 million tons per year of refined metal capacity by financial year 2029 after commissioning its integrated zinc smelter at Dabari. It will also raise silver refining capacity from about 800 to 1500 tons a year. Number 2. A June 2026 study by CEEW and
21:38
Unidentified speaker (speaker_0): NRDC. found that 1 megawatt of rooftop solar creates about 45% years of work. For the same capacity in a large solar park, it's barely 1% a year. Much of that difference comes from selling rather than installing rooftop solar. Business development alone accounts for about 17% years or roughly
21:58
Unidentified speaker (speaker_0): 39% of the total. Only about 13 lakh of the 44 lakh job years. projected by 2030 are longer-lasting operations, maintenance, and manufacturing roles. Number 3. The US Commerce Department has issued an interim rule that would require data center operators in Singapore, Malaysia, Thailand,
22:18
Unidentified speaker (speaker_0): Indonesia, and Brazil to obtain a US government license. The license would be required before exporting advanced AI chips from Nvidia. and other American manufacturers. The rule targets a loophole. Chinese firms are barred from importing top-end American chips, but could until now legally access
22:38
Unidentified speaker (speaker_0): that computing power remotely through data centers in countries such as Singapore, Malaysia, and Thailand. Number 4. Trade between India and Thailand grew 12.6% to $21.49 billion in financial aid. year 2026, according to the Ministry of Commerce and Industry. This followed the 14th India Thailand Joint Trade Committee
22:58
Unidentified speaker (speaker_0): meeting. The two countries discussed improving payment connections through a proposed link between India's UPI and Thailand's Prompt Pay. They also discussed local currency settlement and fintech cooperation between RBI and the Bank of Thailand. But the UPI Prompt Prom Pay connection is still only under discussion. Number 5.
23:17
Unidentified speaker (speaker_0): India is seeking nearly $10 million in investment across five greenfield shipbuilding clusters. The proposed locations are Turkapatnam in Andhrapesh, Kuchadi in Gujarat, Digi in Maharashtra, Kenrapada in Odisha, and Tuthkodi Kodi in Tamil Nadu. India aims to become one of the world's top 10
23:37
Unidentified speaker (speaker_0): shipbuilding nations by 2030. The government has identified domestic demand for more than 400 vessels worth near $24 billion. But that demand still needs to become confirmed orders and operating shipyards. That's all the news I have for you. Thank you so much for watching and see you in the next
23:56
Unidentified speaker (UNKNOWN): one.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.

More from The Daily Brief

The Daily Brief

Indian hotels name their price, and it’s steep

Episode discussion: Indian hotels leverage supply constraints for record profits; hospitals face margin pressure from drug caps.

Speakers’ claims: Speakers note that Indian hotel chains like Leela and IHCL achieved record quarterly profits driven by 8-10% rate hikes rather than occupancy gains, fueled by limited supply and high demand from MICE events and weddings. The industry is shifting toward asset-light management contracts to expand without heavy capital expenditure. In the healthcare sector, hospital chains such as Max Healthcare and Medanta face margin pressure from capped prices on cancer drugs and delayed government payments, prompting a strategic shift toward higher-paying private patients. Additional market tidbits include rising semiconductor costs for automakers, ONGC taking control of an oil block from Vedanta, and a drop in Indian household health spending share.

Indian HotelsHealthcareAsset-Light Models
36 min0 assets10 claims to verify
The Daily Brief

Coal India’s largest subsidiary goes public

Episode discussion: Coal India subsidiary MCL IPO details, regulatory risks, and Indian wire market dynamics discussed

Speakers’ claims: Speakers analyze the upcoming IPO of Mahanadi Coalfields Limited (MCL), noting it is an offer-for-sale transaction where proceeds go to parent Coal India rather than MCL itself. They highlight MCL's operational efficiency and low costs but warn of lower coal quality and significant regulatory risks, including a potential 31,000 crore rupee tax liability in Odisha. The discussion also covers the Indian wire market, where new entrants like Ultratech Cement are challenging incumbents like Polycab and Havells. Incumbents are responding by shifting focus to high-voltage cables and exports to mitigate competition in the wire segment.

IPOCoal IndiaMahanadi Coalfields
27 min0 assets10 claims to verify
All insights